Abstract
A study analyzing a method to detect the leak quantity and its location through computational modeling is presented. Specifically, there is a focus on the transportation of wet natural gas, or methane with a mixture of condensates. Pipeline inclination was an important component of the study as the model considered a pipeline constructed within hilly terrain. Analysis was performed for wet gas pipelines through transient simulations in the OLGA software suite.
Several scenarios were modeled for this study: baseline cases at different operating pressures with no leak, cases with leaks modeled in downhill pipe segments, and cases with leaks modeled in uphill pipe segments. The leaks were modeled with identical characteristics for the two latter cases and were relatively small. It was determined that leak detection is possible in OLGA when trends in certain parameters are monitored. For this study, the minor and moderate leaks were seemingly undetectable using pressure but were detectable using trends in volumetric flow rates at locations upstream and downstream of the leak, and at the pipeline outlet. In addition, the magnitudes of fluctuations in flow parameters were increased significantly at higher pipeline pressures. The severe leak cases were detectable using all parameters.
Introduction
Pipelines are the safest and most economical method for transporting oil and gas across long distances. In order to maintain a high level of safety, leak detection is an important part of daily operations. There are two primary methods of identifying a leak in a petroleum pipeline: through direct assessment or physical inspection (non-continuous method) and model simulation (continuous method).
The United States Department of Transportation (DOT) requires that physical inspections are performed by the operator on a set frequency based on risk assessments and other factors. Despite the regulation requirements, physical inspections can have a negative impact on pipeline economics by creating a need to take segments offline and allocate resources to the work. However, a computer simulation allows the operator to analyze potential leak scenarios and compare them to current pipeline performance while remaining online. The key disadvantages to simulations include the need for prior regulatory approval, and a lack of adequate information for assessing the integrity of the pipeline without concurrent use of other tools. It is worth noting that this study is focused on the continuous leak detection method but does not disregard the regulatory requirement for performing direct assessment. There are various types of leak detection systems (LDS) that are often used concurrently. A brief overview of these non-continuous and continuous LDS will be discussed.
